{"title":"Skill-Driven Data Sampling and Deep Learning Framework for Minute-Scale Solar Forecasting with Sky Images","authors":"Amar Meddahi, Arttu Tuomiranta, Sebastien Guillon","doi":"10.1002/solr.202400664","DOIUrl":null,"url":null,"abstract":"<p>Accurate very short-term solar irradiance forecasting is crucial for optimizing the integration of solar energy into power systems. Herein, an image-based deep learning framework for minute-scale solar irradiance prediction is presented. The locally developed model is benchmarked against two commercial forecasting solutions deployed at the same experimental site, demonstrating superior accuracy and adaptability. A key contribution is the introduction of a skill-driven sampling algorithm based on clear sky index persistence error, which optimizes the training dataset by excluding low-utility samples while retaining essential physical features like solar zenith and azimuth angles. This algorithm enables the exclusion of up to 30% of the original training data, resulting in ≈16% savings in computational resources without affecting forecast accuracy validated using a test set of 324 991 observations. The model achieves a skill score of 7.63%, significantly outperforming the commercial models, which exhibit negative skill scores under the same conditions.</p>","PeriodicalId":230,"journal":{"name":"Solar RRL","volume":"9 4","pages":""},"PeriodicalIF":6.0000,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Solar RRL","FirstCategoryId":"5","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/solr.202400664","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENERGY & FUELS","Score":null,"Total":0}
引用次数: 0
Abstract
Accurate very short-term solar irradiance forecasting is crucial for optimizing the integration of solar energy into power systems. Herein, an image-based deep learning framework for minute-scale solar irradiance prediction is presented. The locally developed model is benchmarked against two commercial forecasting solutions deployed at the same experimental site, demonstrating superior accuracy and adaptability. A key contribution is the introduction of a skill-driven sampling algorithm based on clear sky index persistence error, which optimizes the training dataset by excluding low-utility samples while retaining essential physical features like solar zenith and azimuth angles. This algorithm enables the exclusion of up to 30% of the original training data, resulting in ≈16% savings in computational resources without affecting forecast accuracy validated using a test set of 324 991 observations. The model achieves a skill score of 7.63%, significantly outperforming the commercial models, which exhibit negative skill scores under the same conditions.
Solar RRLPhysics and Astronomy-Atomic and Molecular Physics, and Optics
CiteScore
12.10
自引率
6.30%
发文量
460
期刊介绍:
Solar RRL, formerly known as Rapid Research Letters, has evolved to embrace a broader and more encompassing format. We publish Research Articles and Reviews covering all facets of solar energy conversion. This includes, but is not limited to, photovoltaics and solar cells (both established and emerging systems), as well as the development, characterization, and optimization of materials and devices. Additionally, we cover topics such as photovoltaic modules and systems, their installation and deployment, photocatalysis, solar fuels, photothermal and photoelectrochemical solar energy conversion, energy distribution, grid issues, and other relevant aspects. Join us in exploring the latest advancements in solar energy conversion research.